Despite interdisciplinary approaches and exhaustive use of classical therapeutic procedures, cancers are still among the leading causes of death. More recent therapeutic concepts aim at incorporating the patient's immune system into the overall therapeutic concept by using recombinant tumor vaccines and other specific measures such as antibody therapy. A prerequisite for the success of such a strategy is the recognition of tumor-specific or tumor-associated antigens or epitopes by the patient's immune system whose effector functions are to be interventionally enhanced. Tumor cells biologically differ substantially from their nonmalignant cells of origin. These differences are due to genetic alterations acquired during tumor development and result, inter alia, also in the formation of qualitatively or quantitatively altered molecular structures in the cancer cells. Tumor-associated structures of this kind which are recognized by the specific immune system of the tumor-harboring host are referred to as tumor-associated antigens. The specific recognition of tumor-associated antigens involves cellular and humoral mechanisms which are two functionally interconnected units: CD4+ and CD8+ T lymphocytes recognize the processed antigens presented on the molecules of the MHC (major histocompatibility complex) classes II and I, respectively, while B lymphocytes produce circulating antibody molecules which bind directly to unprocessed antigens. The potential clinical-therapeutical importance of tumor-associated antigens results from the fact that the recognition of antigens on neoplastic cells by the immune system leads to the initiation of cytotoxic effector mechanisms and, in the presence of T helper cells, can cause elimination of the cancer cells (Pardoll, Nat. Med. 4:525-31, 1998). Accordingly, a central aim of tumor immunology is to molecule/1y define these structures. The molecular nature of these antigens has been enigmatic for a long time. Only after development of appropriate cloning techniques has it been possible to screen cDNA expression libraries of tumors systematically for tumor-associated antigens by analyzing the target structures of cytotoxic T lymphocytes (CTL) (van der Bruggen at al., Science 254:1643-7, 1991) or by using circulating autoantibodies (Sahin et al., Curr. Opin. Immunol. 9:709-16, 1997) as probes. To this end, cDNA expression libraries were prepared from fresh tumor tissue and recombinantly expressed as proteins in suitable systems. Immunoeffectors isolated from patients, namely CTL clones with tumor-specific lysis patterns, or circulating autoantibodies were utilized for cloning the respective antigens.
In recent years a multiplicity of antigens have been defined in various neoplasias by these approaches. The class of cancer/testis antigens (CTA) is of great interest here. CTA and genes encoding them (cancer/testis genes or CTG) are defined by their characteristic expression pattern [Tureci et al, Mol Med Today. 3:342-9, 1997]. They are not found in normal tissues, except testis and germ cells, but are expressed in a number of human malignomas, not tumor type-specifically but with different frequency in tumor entities of very different origins (Chen & Old, Cancer J. Sci. Am. 5:16-7, 1999). Serum reactivities against CTA are also not found in healthy controls but only in tumor patients. This class of antigens, in particular owing to its tissue distribution, is particularly valuable for immunotherapeutic projects and is tested in current clinical patient studies (Marchand at al., Int. J. Cancer 80:219-30, 1999; Knuth et al., Cancer Chemother. Pharmacol. 46:p 46-51, 2000).
However, the probes utilized for antigen identification in the classical methods illustrated above are immunoeffectors (circulating autoantibodies or CTL clones) from patients usually having already advanced cancer. A number of data indicate that tumors can lead, for example, to tolerlzation and anergization of T cells and that, during the course of the disease, especially those specificities which could cause effective immune recognition are lost from the immunoeffector repertoire. Current patient studies have not yet produced any solid evidence of a real action of the previously found and utilized tumor-associated antigens. Accordingly, it cannot be ruled out that proteins evoking spontaneous immune responses are the wrong target structures.